| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Article |
From the Divisions of Pediatric Gastroenterology, Hepatology, and
Nutrition,*
Pathology,
and
Developmental Biology,
Children’s Hospital
Research Foundation and Department of Pediatrics, University of
Cincinnati, Cincinnati, Ohio
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plasminogen is a liver-derived serine protease with a well-established role in vascular hemostasis, cell migration, and tissue remodeling through fibrin degradation.6 The genetic inactivation of the plasminogen gene in mice allows for normal development, growth to adulthood, and reproduction.6 However, plasminogen deficiency results in the multi-system accumulation of fibrin, which creates a physical impediment to epithelial cell migration and leads to microvascular thrombosis, delayed tissue repair, and a reduced life expectancy.6-11 Plasminogen is also central to the control of liver repair after an acute injury, but does so in a fashion independent of fibrin degradation. Without plasminogen, restoration of normal hepatic architecture is impaired following an acute toxic injury with persistence of necrotic hepatocytes within injured zones and poor reorganization of hepatic matrix in the immediate reparative process.5 Because of the defective matrix remodeling in these livers, we reasoned that plasminogen-deficient mice could provide a valuable in vivo model to study the role of impaired proteolysis in fibrogenesis. To further explore the regulatory role of plasminogen in remodeling of matrix components in a model of hepatic fibrogenesis, we induced chronic liver injury in plasminogen-deficient mice. Our working hypothesis was that absence of plasminogen leads to the accumulation of matrix components and abnormal lobular reorganization following a chronic liver injury.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Carbon tetrachloride (CCl4) was obtained
from Aldrich (Milwaukee, WI). The sedatives ketamine and xylazine were
obtained from Phoenix Pharmaceuticals (St. Joseph, MO) and acepromazine
maleate was obtained from Fort Dodge Laboratories (Fort Dodge, IA). The
Bio-Rad protein assay was purchased from Bio-Rad (Hercules, CA). Cell
proliferation was determined using the cell proliferaton kit (Amersham
Pharmacia, Piscataway, NJ). The Vectastatin ABC-AP and -HRP detection
systems (Vector Laboratories, Burlingame, CA) and Fast Red
TR/Naphthol AS-MX (Sigma, St. Louis, MO) were used for fibrin
immunohistochemistry. Staining of hepatic stellate cells was performed
using anti-human smooth muscle actin/anti-human smooth muscle
actin-alpha (
-SMA) antibody containing horseradish peroxidase
and anti-human desmin antibody (Dako, Denmark) and M.O.M.
immunodetection kit solution (Vector Laboratories). Gels containing 8%
Tris-glycine acrylamide, 10% gelatin gels, and PVDF membranes
for Western blotting were obtained from Novex (San Diego, CA). A Storm
860 processor (Molecular Dynamics, Inc., Sunnyvale, CA) was used to
detect specific signals from chemifluorescence.
Gene-Targeted Mice
Mice with a targeted disruption of the genes coding for plasminogen (Plg0), fibrinogen (Fib0), or both (Plg0/Fib0) were 1 to 4 months old and of a mixed genetic background and, therefore, all studies were conducted using littermate controls.6,12,13 Polymerase chain reaction-based genotyping was performed as previously described.5 Animal protocols were approved by the Institutional Animal Care and Use Committee of the Children’s Hospital Research Foundation.
Liver Injury
Gene-targeted and control mice were injected biweekly with 0.25 µl of CCl4 per gram body weight as a 12.5% solution in corn oil intraperitoneally.14,15 Mice were examined daily and sacrificed after 1, 4, and 6 weeks of injections. A separate group of Plg0 and Plg+ mice were injected with 0.9% NaCl (saline) in a similar fashion and served as controls. At the time of sacrifice, mice were weighed and anesthesized intramuscularly with 0.1 ml of ketamine:xylazine:acepromazine (4:1:1) per 30 g of body weight. Blood samples were obtained by cannulation of the inferior vena cava. The liver was removed and weighed, and the anterior lobe of the liver was fixed in 10% formalin overnight, followed by paraffin embedding. Determination of serum albumin, alanine aminotransferase (ALT), and bilirubin was performed in plasma using an automated enzymatic assay with the Vistros Chemistry Systems 950.16
Protein and DNA Contents and Cell Proliferation
Frozen liver samples were used for determination of protein (Bio-Rad Assay) and DNA contents.16,17 The proliferative response after CCl4 injury was measured by the incorporation of bromodeoxyuridine (BrdU) into hepatocytes detected immunohistochemically in liver sections as previously described.16 For each liver sample, the hepatocyte-labeling index (percentage of hepatocytes incorporating BrdU) was calculated by counting BrdU-labeled and unlabeled hepatocytes in 10 high-power fields (HPF) (100 hepatocytes nuclei per field) by an investigator unaware of the mouse genotype.16
Fibrinogen and Trichrome Stains
Fibrinogen immunostaining of liver sections was performed using rabbit anti-mouse fibrinogen antiserum, the Vectastatin ABC-AP detection system, and Fast Red TR/Naphthol AS-MX as described previously.6 The standard Mason’s trichrome staining procedure was used to detect ECM components (eg, collagen) in liver sections of control and experimental animals.
Hepatic Stellate Cell Detection
Paraffin-embedded sections of mouse liver were deparaffinized with
xylene and treated with 3% hydrogen peroxide to inhibit endogenous
peroxidase. Following antigen retrieval,16
sections were
initially incubated with blocking reagent. For desmin immunostaining,
anti-human desmin antibody was applied to the sections followed by
biotinylated anti-mouse IgG. Specific signal was obtained with the
Vectastatin ABC reagent and diaminobenzidine substrate, and
counterstaining with Harris’ hematoxylin. The same protocol was
followed for -SMA immunostaining in liver sections, except that
incubation with a secondary antibody was not necessary because
horseradish peroxidase was already present on the antibody. The number
of desmin or -SMA-labeled hepatic stellate cells (HSCs) of each
liver sample was calculated by counting stained HSCs in one high-power
field above and below 10 central veins and portal tracts by an
investigator unaware of the mouse genotype.
Gelatin-Based Zymography and Western Blotting
Gelatin-based zymography was performed as previously described.18 Briefly, liver samples were homogenized and 240 µg of protein lysate from each liver sample was mixed with non-reducing sample buffer and loaded into gelatin-containing polyacrylamide gels. Following electrophoresis, the gels were incubated with 2.5% Triton-X solution twice for a total of 60 minutes and were incubated overnight in 0.1 mol/L glycine solution at 37°C. The gels were then stained with Coomassie blue solution and destained in a solution of 10% acetic acid and 20% methanol to reveal lytic zones. Positive controls consisted of 1 µg of human recombinant MMP-9 (Oncogene Research Products, San Diego, CA). The relative concentration of active MMP-9 in tissue extracts was established by Western analysis as described previously, using antibodies that recognize the active form of the metalloprotease (Oncogene Research Products). Detection of a specific signal was obtained by chemifluorescence (Chemicon International, Temecula CA). To control for loading variability, the relative concentration of albumin was determined using an anti-mouse albumin antibody (Bethyl Laboratories, Montgomery, TX).
Statistical Analysis
Values are shown as means ± standard deviation (S.D.) Statistical significance between two genotypes was assessed by the unpaired t-test. If more than two genotypes were compared, statistical analysis was performed using the General Linear Model and PDIFF option in SAS version 8.01 for Windows. One-way analyses of variance were performed using the number of hepatic stellate cells, liver mass, serum ALT, and percentage of positive-BrDU cells as dependent variables and genotype as the independent variable. The PDIFF option was used to generate pair-wise comparisons among genotypes. A p-value of 0.05 or less was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To examine the impact of plasminogen deficiency on matrix
remodeling following a chronic injury, we injected
CCl4 biweekly into both
Plg+ and Plg0 mice for 1,
4, or 6 weeks. Livers of Plg+ mice had a normal
appearance in the first 4 weeks of injections and exhibited a mild lacy
appearance after 6 weeks. In contrast, livers of
Plg0 mice exhibited a diffusely lacy and pale
appearance at 1 and 4 weeks, becoming enlarged with a grossly pale
micronodular surface with scattered red foci at 6 weeks (Figure 1)
. Saline-treated mice of both genotypes
exhibited a normal hepatic appearance throughout the experiment.
Microscopically, Plg+ livers developed
pericentral necrosis 48 hours after the first dose of
CCl4, and this lesion resolved after the first
week of CCl4 injections.
Plg+ livers subsequently appeared normal through
4 weeks of CCl4; at 6 weeks, minimal accumulation
of eosinophilic material was noted in centrilobular area. A more
pronounced accumulation of this material was already present in the
centrilobular areas of Plg0 livers at 1 week.
After 6 weeks of biweekly CCl4 injections,
Plg0 livers displayed a pattern of pericentral
bridging that, together with persistent necrotic cells, produced the
micronodular/pale appearance, with remaining normal periportal
hepatocytes comprising the red foci (Figure 2)
. Taken together these data show that
livers of Plg+ mice are able to restore and
maintain normal lobular organization until 6 weeks of biweekly
CCl4 injections. In contrast, livers of
Plg0 mice already display a persistent
centrilobular injury following two injections of
CCl4 (first week of injections) in a similar
fashion to the outcome previously reported following an acute
CCl4 insult.5
Thereafter, this
disordered lobular structure progresses in severity over time.
|
|
Because of the enlarged appearance of the Plg0
livers and the increase in eosinophilic material noted microscopically,
liver mass was determined. Liver mass of Plg+
mice increased by 22 ± 13% in mice treated with
CCl4 above saline-treated mice at 6 weeks
(p = 0.002). However, a much more pronounced
increase was observed in livers of Plg0 mice,
which were 77 ± 25% larger than saline-treated
Plg0 livers at one week (p
= 0.007) and reached a maximum increase of 317 ± 72% at 6 weeks
(p < 0.001). This increase was not due to
differences in hepatic DNA content (543 ± 94 µg and 500 ±
146 µg at 6 weeks for Plg+ and
Plg0, respectively, p = 0.46).
Rather, this difference was largely a function of increased total
protein in livers of Plg0 mice (Figure 3A)
. Based on immunohistochemical
analysis, one component of the accumulated protein was fibrin, which is
one established proteolytic target of plasmin (Figure 3B)
. Therefore,
we next determined the mechanistic role of fibrin accumulation in the
inability of Plg0 livers to restore lobular
integrity during chronic injury.
|
To directly determine the role of fibrin(ogen) in the
abnormal repair observed in Plg0 mice following a
chronic injury, we administered CCl4 to mice with
a combined deficiency in plasminogen and fibrinogen. We found an early
development of gross and microscopic injury in
Plg0/Fib0 livers, similar
to that displayed in Plg0 mice. Interestingly,
livers of Fib0 mice developed a mild lacy
appearance by 6 weeks of treatment that was similar to livers of
Plg+/Fib+ mice (Figure 4)
. Microscopically, a considerable
accumulation of eosinophilic material was present in the centrilobular
regions of Plg0/Fib0 at 4
weeks, which progressed to a pattern of pericentral bridging after 6
weeks of CCl4 injections (data not shown),
indistinguishable from that displayed by Plg0
livers after the same length of repeated toxic injury (Figure 2)
. These
data demonstrate that fibrin accumulation in Plg0
livers is not solely responsible for the defect in repair, and that
fibrin deposition is not mechanistically coupled to the abnormal matrix
deposition.
|
. At the ultrastructural level,
Plg+ livers had small amounts of collagen-like
fibrillary material surrounding the central veins, whereas
Plg0 livers had a marked increase in the amount
of fibrillary material in the sinusoids and subendothelial spaces
(Figure 5)
. Taken together, these data identified a prominent
accumulation of hepatic ECM in mice lacking plasminogen. Given that
plasminogen is relatively poor at degrading collagen but a reasonably
effective activator of pro-MMPs in vitro, collagen
accumulation in hepatic tissues may result from a failure of
plasmin-mediated pro-MMP activation.
|
A potential relationship between plasmin and pro-MMP-9 activation
has been proposed in extra-hepatic sites.8
Therefore, we
determined the impact of plasminogen deficiency on hepatic MMP-9
activity in livers collected from Plg+ and
Plg0 mice following chronic injection of
CCl4 or saline. Gelatin-based zymography assays
with liver extracts showed that saline-treated livers contained only
small amounts of gelatinase activity. The amount of gelatinase activity
in the liver increased similarly in mice of both genotypes after 1 and
4 weeks of CCl4 administration. However, after 6
weeks, the gelatinase activity present in liver extracts from
Plg0 mice appeared to be greater than in liver
extracts from Plg+ mice (Figure 6A)
. Using an antibody that specifically
detects only activated MMP-9, Western blot analysis confirmed that
activated MMP-9 was present in the protein lysates of both genotypes
treated with CCl4 (Figure 6B)
. These data suggest
that the accumulation of collagen-like fibrillary material in
Plg0 livers is not due to diminished activation
of MMP-9 in the absence of plasmin, but, rather, to the consequences of
the failure of plasmin in some other as yet unidentified proteolytic
pathways.
|
In addition to impaired proteolysis, centrilobular matrix
accumulation may also result from increased ECM production. Following
an injury, production of fibrillary and non-fibrillary matrix
components derives primarily from activation of HSCs.2,3
To determine whether activation of HSCs may contribute to the excessive
ECM accumulation observed in Plg0 livers, we
determined the number of -SMA-stained HSCs following
CCl4. Plg0 mice treated
with CCl4 had a significantly increased number of
activated HSCs in both periportal and pericentral regions compared to
Plg+ mice treated with CCl4
at 6 weeks (p < 0.004) (Table 1)
. HSCs within the unaffected parenchyma
did not display -SMA staining but could be identified with
anti-desmin antibody, suggesting that activation of HSCs was restricted
to zones of injury (data not shown). Notably, desmin staining also
showed that the total number of HSCs in the zone of injury did not
change between genotypes (Table 1)
. These data suggest that plasminogen
deficiency leads to persistent activation of HSCs within the diseased
microenvironment, which may establish an ongoing state of matrix
production and further impair the defective liver repair of
Plg0 mice.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The prominent and persistent defect in the clearance of necrotic tissues following chronic toxic injury in plasminogen-deficient mice is qualitatively similar to the defect previously reported following an acute injury.5 Specific to chronic injury, however, is the progressive retention of necrotic cells and accumulation of matrix components that result in a greater increase in liver mass. In extra-hepatic tissues, fibrin accumulation is the main mechanism of defective wound healing in the plasminogen-deficient state, and accumulation of fibrin appears to form a physical barrier to the migration of epithelial cells from the wound edges.7,9 The removal of fibrin completely rescued plasminogen-deficient mice from defective wound healing in both skin and cornea.12 Although fibrin is an important component of the provisional hepatic matrix following an injury,19 the findings presented here show that genetically superimposed loss of fibrinogen in plasminogen-deficient mice does not correct the reparative failure in the liver. These data point to the existence of one or more fibrin-unrelated targets for plasmin-mediated proteolysis that are important for hepatic repair within cellular and extracellular matrices.
Plasminogen may be involved in the clearance of cellular and extracellular components either directly or through activation of MMPs or growth factors.20 MMPs are directly involved in matrix degradation and have often been proposed to be linked to the plasminogen activator/plasminogen system of proteases.21,22 Plasminogen has been shown to directly activate MMPs in vitro.4 Impairment in plasminogen activation has been suggested to result in impaired activation of MMP-9 and defective scar formation in a model of acute myocardial infarction in vivo.8 In the liver, we did not find that activation of MMP-9 is impaired in the absence of plasminogen despite the obvious deposition of collagen-like fibrillary material in the areas of abnormal repair. Nevertheless, the available data do not exclude the possibility that plasminogen is required for activation of other pro-MMPs that may be involved in liver regeneration and repair.1,23,24
The prominent activation of collagen-producing HSCs in the livers of CCl4-treated Plg0 mice suggests that production, rather than proteolytic matrix turnover, may be primarily responsible for the observed accumulation of fibrillary material in these mice. In addition to impaired proteolytic removal of ECM components in livers of Plg0 mice, matrix accumulation also results from increased production, as demonstrated by an increase in the activation of HSCs. On injury, HSCs proliferate, display a myofibroblast phenotype, and produce matrix substrates.2,3 This response to injury was not affected by the plasminogen status, but the absence of plasminogen clearly led to a more pronounced activation of HSCs within the injured microenvironment. Although the current studies do not formally determine which molecular factor(s) direct the phenotype of HSCs, one hypothesis consistent with the current findings is that matrix and cellular debris in injured centrilobular areas provide signals that favor activation of HSCs and persistent biosynthesis of collagen-rich matrices. As a consequence, the progressive accumulation of matrix substrates in livers of Plg0 mice results from the combined effects of impaired proteolysis and increased matrix production.
Plasminogen-deficiency provides a valuable in vivo model for studies concerning the role of impaired proteolysis in hepatic matrix production and clearance following chronic liver injury. The molecular mechanisms regulating plasmin-driven matrix proteolysis outside the context of fibrin degradation have not yet been uncovered. Future studies quantifying the impact of plasminogen deficiency on whole tissue proteolysis and the identification of remaining matrix components that trigger and maintain activation of HSCs will be important to define the direct role of plasmin-mediated proteolysis on fibrogenesis. Here, our studies clearly show that plasminogen deficiency leads to a significant accumulation of matrix components after a chronic injury. Since excessive accumulation of matrix elements plays an integral role in this abnormal liver repair, Plg0 mice may be useful in addressing whether therapeutically targeted plasmin-mediated proteolysis can correct alterations in matrix homeostasis.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by National Institutes of Health training grant DK 07727 (J.F.P.), HL 47826 (J.L.D.), and DK 55710 (J.A.B).
Current address for John F. Pohl is the Division of Pediatric Gastroenterology, Scott and White Memorial Hospital, Texas A & M University Health Sciences Center, Temple, TX 76508, and the current address for Hector Melin-Aldana is the Division of Pathology, Children’s Memorial Hospital, Northwestern University, Chicago, IL 60614.
Accepted for publication September 9, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Shanmukhappa, R. Mourya, G. E. Sabla, J. L. Degen, and J. A. Bezerra Hepatic to pancreatic switch defines a role for hemostatic factors in cellular plasticity in mice PNAS, July 19, 2005; 102(29): 10182 - 10187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J G. Duplantier, L Dubuisson, N Senant, G Freyburger, I Laurendeau, J-M Herbert, A Desmouliere, and J Rosenbaum A role for thrombin in liver fibrosis Gut, November 1, 2004; 53(11): 1682 - 1687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Currier, G. Sabla, S. Locaputo, H. Melin-Aldana, J. L. Degen, and J. A. Bezerra Plasminogen directs the pleiotropic effects of uPA in liver injury and repair Am J Physiol Gastrointest Liver Physiol, March 1, 2003; 284(3): G508 - G515. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Yu, F. Wang, C. Jin, X. Wu, W.-k. Chan, and W. L. McKeehan Increased Carbon Tetrachloride-Induced Liver Injury and Fibrosis in FGFR4-Deficient Mice Am. J. Pathol., December 1, 2002; 161(6): 2003 - 2010. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
